Difference between revisions of "Residual statics corrections"

From SEG Wiki
Jump to: navigation, search
(added content)
 
(Velocity analysis and statics corrections: fixed links)
 
(7 intermediate revisions by the same user not shown)
Line 15: Line 15:
 
}}
 
}}
  
 +
==Basic data processing sequence==
 
There is one additional step in conventional processing of land and shallow-water seismic data before stacking — residual statics corrections. From the NMO-corrected gathers in Figure 1.5-26a, note that the events in CMP 216 are not as flat as they are in the other gathers. The moveout in CMP gathers does not always conform to a perfect hyperbolic trajectory. This often is because of near-surface velocity irregularities that cause a static or dynamic distortion problem. Lateral velocity variations caused by a complex overburden can cause move-outs that could be negative — a reflection event arrives on long-offset traces before it arrives on short-offset traces. Close examination of the velocity spectra indicates that some are easier to pick (Figure 1.5-27a) than others (Figure 1.5-28a). The velocity spectrum that corresponds to CMP 297 has sharp coherency peaks that are associated with a distinctive velocity trend. However, the velocity spectrum that corresponds to CMP 188 does not yield a distinctive trend, thus making it relatively difficult to pick (Figure 1.5-28a).
 
There is one additional step in conventional processing of land and shallow-water seismic data before stacking — residual statics corrections. From the NMO-corrected gathers in Figure 1.5-26a, note that the events in CMP 216 are not as flat as they are in the other gathers. The moveout in CMP gathers does not always conform to a perfect hyperbolic trajectory. This often is because of near-surface velocity irregularities that cause a static or dynamic distortion problem. Lateral velocity variations caused by a complex overburden can cause move-outs that could be negative — a reflection event arrives on long-offset traces before it arrives on short-offset traces. Close examination of the velocity spectra indicates that some are easier to pick (Figure 1.5-27a) than others (Figure 1.5-28a). The velocity spectrum that corresponds to CMP 297 has sharp coherency peaks that are associated with a distinctive velocity trend. However, the velocity spectrum that corresponds to CMP 188 does not yield a distinctive trend, thus making it relatively difficult to pick (Figure 1.5-28a).
  
To improve stacking quality, [[residual statics corrections]] are performed on the moveout-corrected CMP gathers. This is done in a surface-consistent manner; that is, time shifts are dependent only on shot and receiver locations, not on the ray paths from shots to receivers. The estimated residual corrections are applied to the original CMP gathers with no NMO correction. Velocity analyses then are often repeated to improve the velocity picks (Figures 1.5-27b and 1.5-28b). With the improved velocity field, the CMP gathers are NMO-corrected (Figure 1.5-26b). Finally, the gathers are stacked as shown in Figure 1.5-29b. For comparison, the stack without the residual statics corrections is shown in Figure 1.5-29a. Reflection continuity over the problem zone between midpoints 53245 has been improved.
+
To improve stacking quality, residual statics corrections are performed on the moveout-corrected CMP gathers. This is done in a surface-consistent manner; that is, time shifts are dependent only on shot and receiver locations, not on the ray paths from shots to receivers. The estimated residual corrections are applied to the original CMP gathers with no NMO correction. Velocity analyses then are often repeated to improve the velocity picks (Figures 1.5-27b and 1.5-28b). With the improved velocity field, the CMP gathers are NMO-corrected (Figure 1.5-26b). Finally, the gathers are stacked as shown in Figure 1.5-29b. For comparison, the stack without the residual statics corrections is shown in Figure 1.5-29a. Reflection continuity over the problem zone between midpoints 53245 has been improved.
 +
 
 +
<gallery>
 +
file:ch01_fig5-26.png|{{figure number|1.5-26}} The NMO-corrected CMP gathers from a land seismic line (a) before and (b) after residual statics corrections. Note that the distorted events (CMP 191, 216) have nearly been flattened (b).
 +
file:ch01_fig5-27.png|{{figure number|1.5-27}} Velocity spectra derived from the same data as in Figure 1.5-26, (a) before and (b) after residual statics corrections. Note that no significant difference exists between spectra derived from CMP gather 297 with and with-out the application of residual statics corrections. Reflection times in this gather did not have significant residual static shifts.
 +
file:ch01_fig5-28.png|{{figure number|1.5-28}} Velocity spectra derived from the same data as in Figure 1.5-26, (a) before and (b) after residual statics corrections. Note the improvement after corrections down to 2.6 s.
 +
file:ch01_fig5-29.png|{{figure number|1.5-29}} The CMP stacks derived from the gathers in Figure 1.5-26. The stack (a) without residual statics corrections shows false structure and poor coherence in the vicinity of CMP 149-197. Both are eliminated by correcting for residual statics (b).
 +
</gallery>
 +
 
 +
==Velocity analysis and statics corrections==
 +
Reflection times often are affected by irregularities in the near-surface. This is best demonstrated by the real data example in Figure 3.3-1. While the shot gathers on the left contain reflections that exhibit nearly hyperbolic moveout, those on the right have reflections that significantly depart from hyperbolic moveout. Although such distortions can be caused by a structural complexity deeper in the subsurface, more often they result from near-surface irregularities.
 +
 
 +
For land data, reflection traveltimes are reduced to a common datum level, which may be flat or vary (floating datum) along the line. Reduction of traveltimes to a datum usually requires correction for the near-surface weathering layer in addition to differences in elevation of source and receiver stations. Estimation and correction for the near-surface effects usually are performed using refracted arrivals associated with the base of the weathering layer ([[refraction statics corrections]]).
 +
 
 +
Traveltime corrections to account for the irregular topography and near-surface weathering layer are commonly known as field statics or [[refraction statics corrections]]. These corrections remove a significant part of the traveltime distortions from the data — specifically, long-wavelength anomalies. Nevertheless, these corrections usually do not account for rapid changes in elevation, the base of weathering, and weathering velocity.
 +
 
 +
Removal of near-surface distortions on reflection times associated with deeper reflectors is routinely done by lowering the shots and receivers along vertical ray-paths from the surface to a datum below the weathering layer. The positioning of shots and receivers to a datum along vertical raypaths amounts to static time corrections in a surface-consistent manner <ref name=ch03r33>Taner et al., 1974, Taner, M. T., Koehler, F., and Alhilali, K. A., 1974, Estimation and correction of near-surface time anomalies: Geophysics, 41, 441–463.</ref>. The term ''static'' implies that it is a constant time shift for an entire trace, and the term ''surface-consistent'' implies that the time correction depends only on the surface location of the shot and receiver associated with the trace.
 +
 
 +
Figures 3.3-2a and 3.3-3a are selected CMP gathers (with [[field statics corrections]]) which were NMO corrected using a set of preliminary velocity picks derived from the velocity analyses in Figure 3.3-4. Deviations from the hyperbolic trends on the CMP gathers significantly degrade the quality of some of the velocity spectra. For instance, [[velocity analysis]] at CMP location 188 yields relatively poorer quality picks than those from other velocity analyses. The CMP gathers in the neighborhood of CMP location 188 have more traveltime distortions compared to some other CMP gathers (Figure 3.3-2a). The resulting stacked section could be misleading in that residual statics may cause dim spots along the reflection horizons as well as false structures (Figure 3.3-5a), particularly between midpoints 101 to 245. False structures also are apparent on the rms AGC gained stack (Figure 3.3-6a) in which dim spots may not be apparent.
 +
 
 +
Obviously a more correct picture of the subsurface should be attained from data corrected for rapidly varying near-surface effects. After making these ''residual'' statics corrections, the CMP gathers with traveltime deviations show better alignment of reflections (Figure 3.3-2b), while those that did not require such corrections are essentially unchanged (Figure 3.3-3b). After the residual statics corrections, the ungained (Figure 3.3-5b) and gained stacked sections (Figure 3.3-6b) show improvement in the continuity of reflections as well as significant elimination of false structures (refer to the segment between midpoints 101 to 245).
 +
 
 +
Following the residual statics corrections, velocity analyses almost always are repeated to update the velocity picks (Figure 3.3-7). Comparison of Figures 3.3-4 and 3.3-7 shows that residual statics corrections have improved the [[velocity analysis]]. The same CMP gathers after NMO corrections using the updated velocity picks are shown in Figures 3.3-8a and 3.3-9a, while the same gathers after residual statics corrections are shown in Figures 3.3-8b and 3.3-9b. Comparison of the CMP gathers before and after residual statics corrections shows significant elimination of time deviations. The resulting stacked sections using the revised velocity estimates are shown in Figure 3.3-10, while the gained stacks are shown in Figure 3.3-11.
 +
 
 +
Compare the stacked sections in Figures 3.3-6a,b and 3.3-11a,b, and observe the gradual improvement in the following order:
 +
 
 +
# CMP stack based on preliminary velocity picks (Figure 3.3-4) but with no residual statics corrections applied (Figure 3.3-6a),
 +
# stack based on preliminary velocity picks (Figure 3.3-4) and with residual statics corrections applied (Figure 3.3-6b),
 +
# stack based on final velocity picks (Figure 3.3-7) but with no residual statics corrections applied (Figure 3.3-11a), and
 +
# stack based on final velocity picks (Figure 3.3-7) and with residual statics corrections applied (Figure 3.3-11b).
 +
 
 +
<gallery>
 +
file:ch03_fig3-1.png|{{figure number|3.3-1}} Common-shot gathers from a land profile. Note the departures from hyperbolic traveltimes on the gathers at the right.
 +
file:ch03_fig3-2.png|{{figure number|3.3-2}} CMP gathers from a land profile: (a) before residual statics corrections, (b) after residual statics corrections. (Shot gathers are shown in Figure 3.3-1.) NMO correction was applied using preliminary velocity picks derived from the spectra in Figure 3.3-4. The CMP stacks are shown in Figure 3.3-5.
 +
file:ch03_fig3-3.png|{{figure number|3.3-3}} CMP gathers from the same land line as in Figure 3.3-2 (a) before and (b) after residual statics corrections. This part of the line does not have as severe a statics problem as that shown in Figure 3.3-2. The NMO correction was applied using preliminary velocity picks derived from the spectra in Figure 3.3-4.
 +
file:ch03_fig3-4.png|{{figure number|3.3-4}} Velocity analyses before residual statics corrections along the land line shown in Figure 3.3-5. Figures 3.3-2 and 3.3-3 show selected CMP gathers.
 +
file:ch03_fig3-5.png|{{figure number|3.3-5}} CMP stacks derived from gathers in Figures 3.3-2 and 3.3-3. Stack (a) before and (b) after residual statics corrections. NMO correction was applied using preliminary velocity picks derived from the spectra in Figure 3.3-4.
 +
file:ch03_fig3-6.png|{{figure number|3.3-6}} The same CMP stacks as in Figure 3.3-5 with rms gain applied. Gained stacks (a) before and (b) after residual statics corrections.
 +
file:ch03_fig3-7.png|{{figure number|3.3-7}} Velocity analyses after residual statics corrections along the line shown in Figure 3.3-5. Figures 3.3-8 and 3.3-9 show selected CMP gathers.
 +
file:ch03_fig3-8.png|{{figure number|3.3-8}} CMP gathers from the land line shown in Figure 3.3-10. CMP gathers (a) without and (b) with residual statics corrections. (Figure 3.3-1 shows the shot gathers from the same line.) NMO correction was applied using final velocity picks derived from the spectra in Figure 3.3-7.
 +
file:ch03_fig3-9.png|{{figure number|3.3-9}} CMP gathers from the same land line as in Figure 3.3-8. CMP gathers (a) without and (b) with residual statics corrections. This part of the line does not have as severe a statics problem as the part in Figure 3.3-8. NMO correction was applied using final velocity picks derived from the spectra in Figure 3.3-7.
 +
file:ch03_fig3-10.png|{{figure number|3.3-10}} CMP stacks derived from the gathers in Figures 3.3-8 and 3.3-9. Stack (a) before and (b) after residual statics corrections and using revised velocity estimates from Figure 3.3-7.
 +
file:ch03_fig3-11.png|{{figure number|3.3-11}} The same CMP stacks as in Figure 3.3-10 with rms gain applied. Gained stacks (a) before and (b) after residual statics corrections and using revised velocity estimates from Figure 3.3-7.
 +
</gallery>
 +
 
 +
Here, the velocity picks made from CMP gathers with no residual statics corrections applied are referred to as ''preliminary'' and those made from CMP gathers with residual statics corrections applied are referred to as ''final''.
 +
 
 +
Residual statics corrections usually are discussed in terms of applications to land data. However, in certain cases, residual statics corrections have produced dramatic improvement in marine data. Areas with irregular water-bottom topography in shallow water (less than 25 m), and areas with rapidly varying velocity in the sediments near the water bottom are places where statics corrections have been successful.
 +
 
 +
Figure 3.3-12 shows a commonly used flowchart for residual statics corrections and [[velocity analysis]] aimed at producing an optimum stacked section. Start with CMP gathers with field statics or [[refraction statics corrections]] applied, and perform [[velocity analysis]], usually no more than a few locations along the line. Then, apply NMO correction with the preliminary velocity picks and compute residual static shifts. Apply these corrections to the original gathers, and repeat the [[velocity analysis]] — this time at all necessary locations along the line. Finally, apply NMO correction and stack the data. In some cases, there may be more than one iteration of estimating and applying the residual statics corrections.
 +
 
 +
In practice, the flowchart in Figure 3.3-12 usually is augmented with additional quality control steps. It often is necessary to examine CMP gathers and velocity analyses after residual statics corrections. Diagnostic tools allow determination of the magnitude of these corrections. For example, common-shot and common-receiver gathers indicate relative static shifts from one receiver location to another and from one shot location to another (Figures 3.3-13 and 3.3-14, respectively). Also, common-shot-point and common-receiver-point stacks can be used in combination with common-receiver and common-shot gathers, respectively. A common-shot-point stack (Figure 3.3-15) should indicate the range of magnitude of shot static shifts; a common-receiver-point stack (Figure 3.3-16) should indicate the range of magnitude of receiver static shifts along the line. These displays enable the determination of an optimum [[maximum allowable shift]] to consider for picking traveltime deviations from the moveout-corrected CMP data for input to residual statics estimation algorithms. From the example in Figures 3.3-15 and 3.3-16, the receiver component of static shifts is greater than the shot component.
 +
 
 +
<gallery>
 +
file:ch03_fig3-12.png|{{figure number|3.3-12}} Processing flowchart with residual statics corrections.
 +
file:ch03_fig3-13.png|{{figure number|3.3-13}} Moveout-corrected common-shot gathers from the same land line as in Figure 3.3-1. Common-shot gathers (a) before and (b) after residual statics corrections.
 +
file:ch03_fig3-14.png|{{figure number|3.3-14}} Moveout-corrected common-receiver gathers from the same land line as in Figure 3.3-13. Common-receiver gathers (a) before and (b) after residual statics corrections.
 +
file:ch03_fig3-15.png|{{figure number|3.3-15}} A diagnostic display for residual statics corrections. Common-shot-point (CSP) stack (a) before and (b) after residual statics corrections. Note the missing shots between CMP 151 and 243. The CSP stack can be used to estimate the magnitude and spatial variation of shot statics along a line.
 +
file:ch03_fig3-16.png|{{figure number|3.3-16}} A diagnostic display for residual statics corrections. Common-receiver-point (CRP) stack (a) before and (b) after residual statics corrections. The CRP stack can be used to estimate the magnitude and spatial variation of receiver statics along a line.
 +
</gallery>
 +
 
 +
==References==
 +
<references/>
  
 
==See also==
 
==See also==
Line 25: Line 88:
 
*[[Velocity analysis]]
 
*[[Velocity analysis]]
 
*[[Normal-moveout correction]]
 
*[[Normal-moveout correction]]
*[[Multiple attenuation]] <!--same page title in Chapter 2-->
+
*[[Multiple attenuation]]
 
*[[Dip-moveout correction]]
 
*[[Dip-moveout correction]]
 
*[[CMP stacking]]
 
*[[CMP stacking]]
Line 32: Line 95:
 
*[[Quality control in processing]]
 
*[[Quality control in processing]]
 
*[[Parsimony in processing]]
 
*[[Parsimony in processing]]
*[[Basic data processing sequence]]
+
*[[Residual statics estimation by traveltime decomposition]]
 
+
*[[Residual statics estimation by stack-power maximization]]
==References==
+
*[[Traveltime decomposition in practice]]
<references/>
+
*[[Maximum allowable shift]]
 +
*[[Correlation window]]
 +
*[[Other considerations]]
 +
*[[Stack-power maximization in practice]]
  
 
==External links==
 
==External links==
Line 41: Line 107:
  
 
[[Category:Fundamentals of Signal Processing]]
 
[[Category:Fundamentals of Signal Processing]]
 +
[[Category:Basic data processing sequence]]
 +
[[Category:Velocity analysis and statics corrections]]

Latest revision as of 08:25, 19 September 2014

Seismic Data Analysis
Seismic-data-analysis.jpg
Series Investigations in Geophysics
Author Öz Yilmaz
DOI http://dx.doi.org/10.1190/1.9781560801580
ISBN ISBN 978-1-56080-094-1
Store SEG Online Store


Basic data processing sequence

There is one additional step in conventional processing of land and shallow-water seismic data before stacking — residual statics corrections. From the NMO-corrected gathers in Figure 1.5-26a, note that the events in CMP 216 are not as flat as they are in the other gathers. The moveout in CMP gathers does not always conform to a perfect hyperbolic trajectory. This often is because of near-surface velocity irregularities that cause a static or dynamic distortion problem. Lateral velocity variations caused by a complex overburden can cause move-outs that could be negative — a reflection event arrives on long-offset traces before it arrives on short-offset traces. Close examination of the velocity spectra indicates that some are easier to pick (Figure 1.5-27a) than others (Figure 1.5-28a). The velocity spectrum that corresponds to CMP 297 has sharp coherency peaks that are associated with a distinctive velocity trend. However, the velocity spectrum that corresponds to CMP 188 does not yield a distinctive trend, thus making it relatively difficult to pick (Figure 1.5-28a).

To improve stacking quality, residual statics corrections are performed on the moveout-corrected CMP gathers. This is done in a surface-consistent manner; that is, time shifts are dependent only on shot and receiver locations, not on the ray paths from shots to receivers. The estimated residual corrections are applied to the original CMP gathers with no NMO correction. Velocity analyses then are often repeated to improve the velocity picks (Figures 1.5-27b and 1.5-28b). With the improved velocity field, the CMP gathers are NMO-corrected (Figure 1.5-26b). Finally, the gathers are stacked as shown in Figure 1.5-29b. For comparison, the stack without the residual statics corrections is shown in Figure 1.5-29a. Reflection continuity over the problem zone between midpoints 53245 has been improved.

Velocity analysis and statics corrections

Reflection times often are affected by irregularities in the near-surface. This is best demonstrated by the real data example in Figure 3.3-1. While the shot gathers on the left contain reflections that exhibit nearly hyperbolic moveout, those on the right have reflections that significantly depart from hyperbolic moveout. Although such distortions can be caused by a structural complexity deeper in the subsurface, more often they result from near-surface irregularities.

For land data, reflection traveltimes are reduced to a common datum level, which may be flat or vary (floating datum) along the line. Reduction of traveltimes to a datum usually requires correction for the near-surface weathering layer in addition to differences in elevation of source and receiver stations. Estimation and correction for the near-surface effects usually are performed using refracted arrivals associated with the base of the weathering layer (refraction statics corrections).

Traveltime corrections to account for the irregular topography and near-surface weathering layer are commonly known as field statics or refraction statics corrections. These corrections remove a significant part of the traveltime distortions from the data — specifically, long-wavelength anomalies. Nevertheless, these corrections usually do not account for rapid changes in elevation, the base of weathering, and weathering velocity.

Removal of near-surface distortions on reflection times associated with deeper reflectors is routinely done by lowering the shots and receivers along vertical ray-paths from the surface to a datum below the weathering layer. The positioning of shots and receivers to a datum along vertical raypaths amounts to static time corrections in a surface-consistent manner [1]. The term static implies that it is a constant time shift for an entire trace, and the term surface-consistent implies that the time correction depends only on the surface location of the shot and receiver associated with the trace.

Figures 3.3-2a and 3.3-3a are selected CMP gathers (with field statics corrections) which were NMO corrected using a set of preliminary velocity picks derived from the velocity analyses in Figure 3.3-4. Deviations from the hyperbolic trends on the CMP gathers significantly degrade the quality of some of the velocity spectra. For instance, velocity analysis at CMP location 188 yields relatively poorer quality picks than those from other velocity analyses. The CMP gathers in the neighborhood of CMP location 188 have more traveltime distortions compared to some other CMP gathers (Figure 3.3-2a). The resulting stacked section could be misleading in that residual statics may cause dim spots along the reflection horizons as well as false structures (Figure 3.3-5a), particularly between midpoints 101 to 245. False structures also are apparent on the rms AGC gained stack (Figure 3.3-6a) in which dim spots may not be apparent.

Obviously a more correct picture of the subsurface should be attained from data corrected for rapidly varying near-surface effects. After making these residual statics corrections, the CMP gathers with traveltime deviations show better alignment of reflections (Figure 3.3-2b), while those that did not require such corrections are essentially unchanged (Figure 3.3-3b). After the residual statics corrections, the ungained (Figure 3.3-5b) and gained stacked sections (Figure 3.3-6b) show improvement in the continuity of reflections as well as significant elimination of false structures (refer to the segment between midpoints 101 to 245).

Following the residual statics corrections, velocity analyses almost always are repeated to update the velocity picks (Figure 3.3-7). Comparison of Figures 3.3-4 and 3.3-7 shows that residual statics corrections have improved the velocity analysis. The same CMP gathers after NMO corrections using the updated velocity picks are shown in Figures 3.3-8a and 3.3-9a, while the same gathers after residual statics corrections are shown in Figures 3.3-8b and 3.3-9b. Comparison of the CMP gathers before and after residual statics corrections shows significant elimination of time deviations. The resulting stacked sections using the revised velocity estimates are shown in Figure 3.3-10, while the gained stacks are shown in Figure 3.3-11.

Compare the stacked sections in Figures 3.3-6a,b and 3.3-11a,b, and observe the gradual improvement in the following order:

  1. CMP stack based on preliminary velocity picks (Figure 3.3-4) but with no residual statics corrections applied (Figure 3.3-6a),
  2. stack based on preliminary velocity picks (Figure 3.3-4) and with residual statics corrections applied (Figure 3.3-6b),
  3. stack based on final velocity picks (Figure 3.3-7) but with no residual statics corrections applied (Figure 3.3-11a), and
  4. stack based on final velocity picks (Figure 3.3-7) and with residual statics corrections applied (Figure 3.3-11b).

Here, the velocity picks made from CMP gathers with no residual statics corrections applied are referred to as preliminary and those made from CMP gathers with residual statics corrections applied are referred to as final.

Residual statics corrections usually are discussed in terms of applications to land data. However, in certain cases, residual statics corrections have produced dramatic improvement in marine data. Areas with irregular water-bottom topography in shallow water (less than 25 m), and areas with rapidly varying velocity in the sediments near the water bottom are places where statics corrections have been successful.

Figure 3.3-12 shows a commonly used flowchart for residual statics corrections and velocity analysis aimed at producing an optimum stacked section. Start with CMP gathers with field statics or refraction statics corrections applied, and perform velocity analysis, usually no more than a few locations along the line. Then, apply NMO correction with the preliminary velocity picks and compute residual static shifts. Apply these corrections to the original gathers, and repeat the velocity analysis — this time at all necessary locations along the line. Finally, apply NMO correction and stack the data. In some cases, there may be more than one iteration of estimating and applying the residual statics corrections.

In practice, the flowchart in Figure 3.3-12 usually is augmented with additional quality control steps. It often is necessary to examine CMP gathers and velocity analyses after residual statics corrections. Diagnostic tools allow determination of the magnitude of these corrections. For example, common-shot and common-receiver gathers indicate relative static shifts from one receiver location to another and from one shot location to another (Figures 3.3-13 and 3.3-14, respectively). Also, common-shot-point and common-receiver-point stacks can be used in combination with common-receiver and common-shot gathers, respectively. A common-shot-point stack (Figure 3.3-15) should indicate the range of magnitude of shot static shifts; a common-receiver-point stack (Figure 3.3-16) should indicate the range of magnitude of receiver static shifts along the line. These displays enable the determination of an optimum maximum allowable shift to consider for picking traveltime deviations from the moveout-corrected CMP data for input to residual statics estimation algorithms. From the example in Figures 3.3-15 and 3.3-16, the receiver component of static shifts is greater than the shot component.

References

  1. Taner et al., 1974, Taner, M. T., Koehler, F., and Alhilali, K. A., 1974, Estimation and correction of near-surface time anomalies: Geophysics, 41, 441–463.

See also

External links

find literature about
Residual statics corrections
SEG button search.png Datapages button.png GeoScienceWorld button.png OnePetro button.png Schlumberger button.png Google button.png AGI button.png