CHDK Wiki
Advertisement

Since the discovery of many new override features, some far surpassing what the cameras were originally sold to do, it has been found that each make and model of camera may have variations in the absolute limits of Shutter-Speeds, F-Stops, ISO-Values, Motion-Detection Speed, among others.

It might be interesting if people who have tested the limits of their particular cameras would provide what limits they found. This would also be a nice quick & dirty chart for anyone wanting to choose a CHDK-capable camera based on these new capabilities. As well as a place to show to newcomers the answer to their "I set my camera to 2 minutes exposure but I only get 65 seconds, Why" questions.

NOTE: All ratings in this chart should be taken as a basic guideline. These were not tested in labs and we have to depend on the creative ways and know-how of each individual camera user that might have tested them. Also, due to discrepancies in build-quality from camera to camera, you may find that your particular camera may not attain or might even surpass the ratings that others might have listed here.


Important! -- Just because you can set an override shutter-speed, f/stop or ISO sensitivity on your camera with CHDK, it doesn't mean your camera can actually do that setting. Be sure you have tested to make sure that extreme setting is actually making a difference in your resulting images. Listing here the available numbers that are built into CHDK will do no good. Those numbers can go well outside of your camera's capabilities. You might select an f/stop of f/16, but it doesn't mean you are actually obtaining that desired setting. EXIF will even record your chosen setting, but it may not actually be applying it to your image. Find ways to test them, to be sure you are actually getting those limits. This list should be a reference of TRUE limits of your cameras, not the possible ones that are only written into CHDK.


New Camera Capabilities with CHDK[]

Values in (parenthesis) are the camera's OEM value.


Name Longest Tv Fastest Tv & Flash-Sync Highest Av ISO Lowest / Highest Fastest Motion-Detection Response 1 RAW File/Frame Sizes 2 Min. USB-Remote Volts / Wakes up on USB signal 9Bayonet / thread (mm)
A450 {} {} {} 75(80) / 800(400) {} 6,553,440 bytes {} / Wake-up: {} {}
A460 {} {} {} 32(80) / 800(400) {} 6,553,440 bytes {} / Wake-up: {} {}
A470 64" (15")

1/2000


{}

{} {} {} 9,243,240 bytes 3.4V NOT OK, 3.5V to 5.0V OK / Wake-up: NO {}
A480

512"

{} {} {} 127ms 15,467,760 bytes

3.0V NOT OK, 3.5 OK, 4.5VOK, 5V OK


Wake-up / {}

{}
A490/A495 80/1600 3.6V OK, 5V OK
A530 2147" / (15")

1/8000


(1/2000)


@F5.6

{} 40(80) / 800(800)

90ms / 130 -220 ms typical


(MDBF-080914)

6,573,120 bytes (JPG 2592 x 1944) 3V OK / Wake-up: {} None
A540 2147" / (15") around 1/30000s @F2.6 35mm f/18 @ 140mm (f/8); f/8 @ 35mm (f/8) {80}(80) / {800}(800) {} 7,710,960 bytes / 2832 x 2120 (JPG 2816 x 2112) 3V OK / Wake-up: {} Bayonet
A550 2147" / (15") {} {} 80/800 {} 9,219,600 bytes / Res. 3112 x 2328 (JPG 3072 x 2304) 3V not OK, 3.7V and 4.5V OK / Wake-up: {} {}
A560 64" (15") {} {} {} / f/8.0 simulated /w ND Filter {} 9,219,600 bytes 3.5V, 4.8V ok / Wake-up: NO {}
A570 IS 2147" / (15") 3/100,000" @ f16/140mm f/16 @ 140mm EFL; f/11 @ 35mm EFL (f/8) 50/5000 60ms / 80-120ms typical (MDFB-080914) 9,219,600 bytes 3V NOT OK,4.5V and 4.8V OK / Wake-up: NO Bayonet
A590 IS 2147" / (15") 3/100,000" @ f16/140mm f/20 @ 140mm EFL; f/11 @ 35mm EFL (f/8) 65(80) / 1600 ; 2500 in P-mode and AutoISO HI (1600) 120 - 150ms with MLuna's script - based on MX395 - 50ms best / 60-90ms typical with fudgey´s MDFB-080914 on allbest-latest Build 10,341,600 bytes / Res. 3296x2472 (3264x2448) 3V NOT OK, 3.6V and 4.5V OK / Wake-up: NO Bayonet
A610 64" (15") 1/30,000" @ f/8-f/11; 1/6,000" @ f/2.8 (1/2,500") f/16 @ 140mm EFL; f/11 @ 35mm EFL (f/8) 30(50) / 600(400) {} 6,573,120 bytes / Res. 4 2616 x 1960 (JPG 2592 x 1944) 3V OK / Wake-up: {} Bayonet
A620 64" (15") 1/33,000" @ f/8-f16 (1/2,500"); 1/16,666" @ f2.8 (1/1,250") f/16 @ 140mm EFL; f/11 @ 35mm EFL (f/8) 26 (50) / 600 in Custom Auto ISO (400) ; 400 (400) remains the limit in bracketing or over-ride modes. Tested w/ CHDK 0.9.7-725 {} 9,219,600 bytes / Res. 4 3112 x 2328 (JPG 3072 x 2304) 3v / Wake-up: {} Bayonet
A630 64"-65" (15") {} f/16 @ 140mm EFL; f/11 @ 35mm EFL (f/8) 80 / 800 {} 10,383,120 bytes / Res. 4 3288 x 2472 (JPG 3264 x 2448)

3.7V OK


3V OK / Wake-up: {}

Bayonet
A640 65" 1/40,000 f/16 @ 140mm EFL; f/11 @ 35mm EFL (f/8) {} / {} {} 12,945,240 bytes / Res. 4 3672 x 2760 (JPG 3648 x 2736) 1.15V / Wake-up: No 6 Bayonet
A650 IS 65" (15") {} f/16 @ 210mm EFL; f/11 @ 35mm EFL (f/8); EXIF data is correct f/2.8-f/11 50(80) / {}(1600)(3200 in custom mode); EXIF data records ISOx1.5 150-200 ms with MDFB-080914 15,636,240 bytes / Res. 4032x3024 (JPG 4000x3000) 3V NOT OK, 4.5V and 4.8V OK / Wake-up: {} Bayonet
A700 64" (15") 1/20,000" @ f/8 F/12.5 @ 210mm EFL {} / Wake-up: {} {} 7,710,960 bytes / Res. 4 2840 x 2128 (JPG 2816 x 2112) {} / {} Bayonet
A710 IS 2147" / (15") {} f/14.25 @ 210mm EFL (f/8) 50 (80) / 800 (800) {} 9,219,600 bytes / Res. 4 (JPG 3072 x 2304) 2.86 V OK / Wake-up: No Bayonet
A720 IS 64"-65" (15") {} f/14 @ 210mm EFL / f/8 @ 35mm EFL (f/8) 40 (80) / 25007 (1600) 75ms / 90-110ms typical (MDFB-080914) 10,341,600 bytes / Res. 4 3298 x 2472 (JPG 3264 x 2448) ~4,5 V OK/ Wake-up: No Bayonet

A800

1/100k 15,106 bytes

? V OK / Wake-up: No

A810 123" (15")

1/100k (1/2000)

(100) / 10000 (1600) 150-200ms ~4.5V - 5.0V OK / Wake-up: No
A1100 IS 15" (15") {} {} 10 (80) / 8000 (1600) {} {} 4.5V-4.6V OK / Wake-up: {} {}
A2000 IS {} {} {} {} ({}) / 8000 (1600) {} {} {}

{}

D10 {} {} {} {} {} {} {} {}
G7 64" {15"} 1/2500 (1/2000) f/11 (f/8) 80 / {} 95ms average with MD_Testv3 12,945,240 bytes / Res. 4 3672 x 2760 (JPG 3648 x 2736) {} / Wake-up: {} {}
G9 64"{15"} 1/2500{1/2500} F/11{F/8} 10{80}/8000{1600(Possible 3200 in preset mode)} 32ms possible / 50-90ms typical {} 3V NOT OK,4.5V and 4.8V OK {}
S2 IS 64"-65" (15") {} {} 10(50) / 409(400) 60-90ms typical with MDFB-080914 6,573,120 bytes / Res. 4 2616 x 1960 (JPG 2592 x 1944) {} / Wake-up: Yes Bayonet 54mm
S3 IS 2147" / (15") 1/40,000" / Flash-Sync 1/60,000" (OEM 1/3,200" & OEM flash-sync to 1/500" only) f/11 for 300mm and longer EFL; f/8 @ 36mm EFL (f/8) 50 (80) / 1000 (800) 45ms sometimes / 60-90ms typical 7,710,960 bytes (7.35MB) / Res: 2840 x 2128 (JPG 2816 x 2112) 3V OK / Wake-up: Yes Bayonet
S5 IS 2147" / (15") 1/33,333" @ f/2.7-f/11 / Flash sync 1/224,000" + Shoe sync (OEM 1/3,200" @ f/8.0 & flash sync up to 1/500" or 1/3,200" (ext.flash, highspeed mode)) 5 f/11 for 300mm and longer; f/8 for 36mm (f/8) 80 / 1600 {} 10,341,600 bytes / Res. 4 3298 x 2472 (JPG 3264 x 2448) 3.7V OK / Wake-up: No fixed
SD400 (IXUS50) {} {} {} {} / {} {} {} 2.15V OK / Wake-up: No {}
SD450 (IXUS 55) 64" {} {} 10 / 429 {} CRW 6,611,520 bytes / JPEG(2592 x 1945) / Frames (na) {} {}
SD500 (IXUS700) {} {} {} {} / {} {} 9,219,600 bytes / Res. 4 (JPG 3072 x 2304) {} / Wake-up: {} {}
SD550 (IXUS750) 64" / (15") {} {} {} / {} {} 9,219,600 bytes / Res. 4 (JPG 3072 x 2304) {} / Wake-up: {} {}
SD600 (IXUS60) {} {} No overrides, aperture is connected to focal length. So F/2.8 for 35mm and F/4.9 for 105mm {} / {} {} 7,710,960 bytes for 1.00a; 7,749,360 bytes for 1.00d / Res. 4 (RAW 2840 x 2128; JPG 2816 x 2112) {} / Wake-up: {} {}
SD630 (IXUS65) {} {} {} {} / {} {} 7,710,960 bytes / Res. 4 (RAW 2832 x 2120; JPG 2816 x 2112) {} / Wake-up: {} {}
SD700 (IXUS800 IS) {} {} {} {} / {} {} 7,710,960 bytes {} / Wake-up: {} {}
SD750 (IXUS75) {} {} {} 50 (80) / 1000 (1600) {} 9,219,600 bytes 3.0V - 3.2V NOT OK, 4.5V OK / Wake-up: no {}
SD780 (IXUS100) {} {} {} {} / {} < 720 ms {} {} / Wake-up: {} {}
SD800 (IXUS850) {} {} {} {} / {} 80-120 ms 9,219,600 Bytes {} / Wake-up: {} {}
SD900 (IXUS900Ti) {} {} {} {} / {} {} 12,945,240 bytes 3,5V / Wake-up: {} {}
SD850 IS (IXUS950 IS) 2147" / (15") 1/5,000 (1/1,600) phys. 2.8-5.5 (linked to zoom), eqiv. 8.0-16.0 with ND8 {} (80) / {} (1600) 100-120ms typically, with luck its <70ms 10,341,600 bytes 3.2 to 3.4V / Wake-up: No {}
SD870 (IXUS860) {} {} {} {} / {} {} 10,341,600 bytes {} / Wake-up: No {}
SD880 (IXUS870) {} {} {} 80 (80) / 800 (1600) 8 {} 15,467,760 bytes (Canon RAW); 15,506,160 bytes (DNG) / 3668 x 2756 (JPG 3648 x 2736) min 3.50V / Wake-up: {} {}
SD950 (IXUS960) {} {} {} {} / {} {} 15,636,240 bytes {} / Wake-up: {} {}
SD890 (IXUS970IS) {} 1/1,600 - {} {} {} 3V ok {}
SD1000 (IXUS70) 2147" / (15") 1/10,000 (1/1,500) f/8.0 simulated /w ND Filter (f/8.0) 60 {80} / 6400 (1600) 160ms / 250-300ms typical 9,219,600 bytes / Res. 4 (JPG 3072 x 2304) min 3,4V / Wake-up: No {}
SD1100 (IXUS80) 2147" / (15") (1/1,500) f/8.0 simulated /w ND Filter (f/8.0) 50 {80} / 6400 (1600) >3.4V (4.5V ok) / Wake-up: No[1] {}
SD4000 (IXUS300) min. 3.4V
SX100 IS 64" (15") 1/40,000 at f/8 (f/11) {} 49 (80)/ 1600 (1600) {} 10,341,600 bytes / min. 3,4V / Wake-up: No {}
SX110 IS 64" (15") {} f/16 (f/8) 50 (80) / 800 (1600) 50ms / 60-90ms typical (with MDFB-080914) 15,467,760 bytes / Res. 3720 x 2772 (JPG 3456 x 2592) 3,6V OK / Wake-up: No {}
SX120 IS 2048" (15") {} f/11 (f/8) 50 (80) / 800 (1600) ? CRW: 15,534,576 bytes, DNG: 15,572,976 bytes / Res. CRW: 3664 x 2742, DNG: 3652 x 2758 (JPG 3648 x 2736) ? {}
SX10 IS 2147" / (15") {} f/11{8} {10} (80) / {8000} (1600) 75 ms typical (with MDFB-080914) 15,467,760 bytes / Res. 3720 x 2772 (JPG 3648 x 2736) 3V NOT OK, 4.5V OK / Wake-up: No Bayonet for lens hood, adapter to 58mm thread exists
SX1 >100" (1") {} {} 80 (80) / 3200 (1600) 100-80 ms (with MDFB-080914) 17,264,016 bytes (15,783,794 bytes) / {} / Wake-up: {} {}
SX200 IS 1024"/ {15"} {}[2] f/16 @ 336mm (?) @ f/11.31 @ 28mm (?) (ND filter?) 7 (10)/ 4700 (8000) {} / Res. (JPG 4000x3000) 3.0V - 3.2V NOT OK, 4.7V OK / Wake-up: {}

{}

SX20 IS {} {} {} (80)/(1800 & 3200) {} {} min. 3.4V / Wake-up: No {}
SX230 HS >16000"(15") {} f/16 (f/8) (DOF-method) 80? (80) / ~5000? (3200) {} 19,131,120 bytes {} {}
SX30 IS (15") (1/25,000)

F/11@24mm


F/16@840mm (F/8)

(80)/(4000)() 21,974,496 bytes / Res. 4352 x 3238 (JPG 4320 x 3240)
SX510 HS 60" (15") 1/1600 (1/1600) F/11 (F/8) 80 (80) / 3200 (3200) {} {} {} {}
S80 64" (15") 1/33000 (1/2000) F/11 (F/8) (50)/(400) is still the limit 80-110ms 10,383,360 bytes 3V OK {}
S90 2048" (15") (1/1600) (F/8) {} {} {} {} {}
IXUS230 HS {} {} Fixed aperture linked to zoom (f/3.0-f/5.9) {} {} 19,169,520 bytes / Res. 4064x3028 (JPG 4000x3000) {} {}

1 RAW File/Frame Sizes - You might notice something you haven't noticed before. In most cameras the RAW image dimensions (not just file-size) are larger than the resulting JPG file dimensions that the camera creates. This is due to the need to throw away pixels around the border, pixels that contain real image data (just beyond your normal viewfinder's FOV). In-camera RAW interpolation routines are not very "intelligent" and don't know how to deal with these border pixels. More advanced editing software usually doesn't have this problem. For example: If you look at the S3 IS data here, you'll see that the RAW image dimensions are 2840 x 2128 and the JPG is 2816 x 2112. That's an extra 24 pixels horizontally and 16 pixels vertically in the RAW image. If the JPG is defining a 36mm (35mm eq.) lens FOV then the RAW is showing you a 35.7mm (35mm eq.) FOV. You might be getting almost an extra 1/3 to 1/2mm focal length of extra wide-angle image data for free with the CHDK RAW image. (Another small perk to shooting with RAW in these cameras.)

2 Min. volts for USB Remote - It has been found that some camera models are more sensitive on what minimum voltage must be used to get the USB-Remote features to trigger cleanly and properly. Not too long ago most cameras that were supported with USB-Remote feature only need a 3V button-cell for the trigger. Today this is no longer true and 3 stacked 1.5v cells (4.5v) might be needed for some cameras. This voltage information would probably be better in some other page in the Wikia, like the Scripting Tutorial or other places that USB-Remote is explained, but for data-collecting purposes this chart might be a good place to collect this info (for now).

Wakes up on USB signal means whether the off-state camera is switching on when a USB signal is detected.

4 RAW Image Dimension derived from DCRAW source-code, JPG Image Dimension derived from DNG4PS source-code.

If when checking your RAW file-sizes for this chart, you might want to consider adding a sample RAW image to this online archive http://www.rawsamples.ch/ Developers of programs have started using this as a resource for samples from many kinds of cameras. If they have samples of these non-standard CHDK RAW formats, more editing and viewing programs might support them one day.

5 Although it seems strange, a difference is still visible between 1/25,000 and 1/33,333 at f/2.7 on the S5IS. That particular test was conducted with a 'non-moving' object (the sun). Later tests, however at f/8 and with full flash, show a 6-degree rotation on a dremel disk turning at 33,000rpm. A quick calculation shows 1/32,520 for the shutter time. At minimal flash, the disk is rotated by just 0.9 degrees, which means a 1/224,000 shutter/flash time. Someone might want to verify these values some time soon, as they look pretty extreme.

By the way, when using the Speedlite 580EX II, OEM flash sync will work all the way to 1/3,200", when the flash is set to highspeed mode. Regardless, the external flash syncs even at 1/33,333" with 1/128 strength. Setting the 580EX II to 1/1 does not generate the expected amount of extra exposure, the flash unit probably requires more than 1/33,333" to reach full power. A quick test (though on a 160cm (5 1/3 ft) offshoe cord) shows that the camera syncs with the flash on 1/33,333", though exposure seems to stop increasing around or just slightly weaker than 1/16. It still syncs at 1/1, but the exposure is the same as 1/16. Also, the exposure at these intensities differs in shoots with the same settings, so it's likely to assume that the sync isn't always perfect and probably lags a bit. For the record: the Speedlite had to recycle a lot longer than normal, so the batteries were probably almost dead. While this does not affect flash intensity, it may cause the internal circuitry to run a bit slower and throw off these measurements. It also seems that ETTL works at this shutter speed. Judging by the sound the flash makes, the second flash is more intense than the preflash and does register on the image (manual flash set to the estimateed preflash strength gives less exposure). This is hard to test and judge, though, I might be wrong.

6 1.15V Seems suspiciously low compared to other cameras, but that's what I measured using my multimeter and an adjustable power supply. I'm using 2xAA battery (3V) with no problems.

7 2500 may not be achievable, to be sure it would need further testing. 1600 seems to be the maximum by some recent (non scientific) testing. An0n: 01/08/09.

8 Values outside these limits can be set in CHDK and are taken into account by the automatic exposure system, but in reality, ISO stays fixed at 80/800, resulting in overexposed/underexposed images. ISO values set in CHDK need to be multiplied by approx. 1.5 to correspond to those reported by the camera on the display and in the EXIF data (reported ISO 800 ˜ CHDK ISO 530). The camera's physical sensitivity is limited to ISO 800; at ISO settings above 800, the camera automatically brightens up the image by the corresponding EV value. This automatic brightening-up is for some reason disabled when the sensitivity is set by CHDK (either via manual ISO override or via Custom auto ISO function), so JPEGs with ISO above 800 come out underexposed (raw files above ISO 800 are always underexposed, regardless of whether the ISO was set by CHDK or the camera).

9 Data for Bayonet or thread to attach lenses as of 4th Nov. 2009 are from: http://www.digitaltoyshop.co.uk/product.asp?p_id=5066&pt_id=932&lg=1&c=CANON_ADAPTER_TUBE_ Please add you finding so one can choose a camera not just by firmware capabilities, but also by way of connecting converter lenses.

Contributors: When adding your own findings to this chart, simply click on the [edit] link just above and to the right of the chart. In your editing window scroll down to your camera model, and insert the information in this sequence:

! Camera Model 
! Lowest Tv (OEM value in parens)
! Highest Tv & Flash-Sync (OEM value in parens)
! Highest Av (OEM value in parens)
! ISO Lowest / Highest (OEM value in parens)
! Fastest Motion-Detection Response
! RAW File/Frame Sizes (OEM values in parens)
! Minimum USB-Remote Voltage / Camera wakes up on USB signal? (Yes No)

How To Test Your Camera[]

Longest Tv[]

How to measure longest Tv?


High Shutter-Speed (Tv) and Flash Speed Tests[]

NOTE: Absolute fastest shutter-speeds available only with highest f/stops.

Due to these new shutter speeds being much faster than any consumer-grade cameras on the market, some rather inventive ways had to be found to determine their true top limits. Please read the following pages and discussions just to see what we were up against in trying to do so, therein you'll find the best ways to test your own camera.

High-Speed Shutter & Flash-Sync Samples
Use a hobby-drill (dremel drill) to test the flash speed. Add a small mirror and laser-pointer to the drill to test your true CCD shutter speed.
CHDK Forum Discussion on myriad ways used to try to find highest shutter speeds.
And these related threads on the CHDK Forum: "Shutter Speed Tests", "High Speed Shutter Anomalies", and "An Ultra High-Speed Shutter Override" (tests by photographing CRT & TV scan lines, an easy-to-do test for slower-shutter cameras).


Highest F/Stop (Av) Tests[]

NOTE: Av override values are only available at long-zoom settings.

Put your lens set to the longest focal-length (longest zoom) and then (using the method described below by Allbest for ISO tests), use Av bracketing mode.
1. Go to menu Extra Photo Operations > Override Aperture > Set to any value at or below your camera's smallest (highest number) aperture. Ex: if your camera's highest f/stop is f/8.0, set it to some value at or just below that. (This way you have a known reference frame to start with.) OR, put your camera in Av Priority Mode, and set it for a small aperture.
2. Go to menu Bracketing in Continuous Mode > Av Bracketing Value > set 1/3EV
3. Bracketing Type > "-"
4. Put camera in continuous shooting mode.
5. Fire off a series of shots to be sure you have obtained images at all override apertures of f/8.0 to f/16 (seven 1/3EV steps from f/8.0 to /16).
6. Analyze histogram shifts in images to see where there is no difference between shots. The last image to show a change in exposure level is your Av limit.
7. Analyze changes of DOF in images to see where there is no difference between shots. The last image to show a change of DOF is your Av limit.


Low/High ISO Tests[]

Testing advice borrowed from a recent post by Allbest:
"You can more correctly test hardware limits of your camera in such sequence of actions:
1.Set ISO mode to AUTO
2.Go to menu entry "Bracketing by ISO".
3.Set bracketing direction to "+".
4.Set some ISO shift. For Example: 20
5.Choose some initial value for ISO via menu entry "Iso override". For example: 10
6.Set continuous shooting mode
7.Capture set of shots
8.Analyze histogram shift at each step and fix moment when there is no differences between shots
9.See Exif info for extreme points you receive via Exiftool (here you'll see marketed values)
10.Change bracketing direction to "-" and repeat 6-9 to test your camera's lowest ISO setting"

Motion-Detection Speed Tests[]

CHDK Forum Discussion "Motion Detection Too Slow" That discussion thread includes some new high-speed scripts and a testing program "MD-Test v2" (by "jonnythe) whereby you can test your camera's motion-detection response time just by photographing your computer screen. (read thread for info on use and optimum md-detection script settings)
Advertisement