Archive 2006 · Contax N to Canon conversion - initial results
brainiac Offline [X]
p.8 #1 · Contax N to Canon conversion - initial results
Bo-ming of Conurus has sent me his address, so the non-functioning unit will be winging its way back to him shortly. Bo-ming is generally very helpful and responsive.
I have also received the replacement circuit which enabled the camera to display an aperture for a second and then the same fault appeared on both cameras, so it looks like this is an electronics fault, i.e. the reverse engineering of the camera to lens instruction has been unsuccessful for 5D/24-85 at least. Bo-ming is working on fixing it and has promised to send me a third circuit once he has ironed out the problem.
My fingers are crossed in the hope that Bo-ming will prevail with what is clearly a sticky problem. I can't wait to get this lens working, but for the moment I have a Contax 28-85 and Canon 28-70 as alternatives, so it's not a huge hardship...
p.8 #2 · Contax N to Canon conversion - initial results
Yep, I am working on the firmware and found a bug in handling shut-down, which sometimes causes intermittent problems.
At least two other customers have successfully deployed the 5D/24-85 combination, so I thought I sent you a defective product, but it is unlikely that both mounts I sent to you stopped working in exactly the same way. Also you have two 5D's from what I have heard, and neither worked.
Could it be that some 5D's are different from the others?
Sep 18, 2006 at 10:34 AM
brainiac Offline [X]
p.8 #3 · Contax N to Canon conversion - initial results
Could it be that some 5D's are different from the others?
Well, I am running firwares 1.03 in one and 1.05 in the other. One is serial 0430100817 and the other is 1030705329.
p.8 #4 · Contax N to Canon conversion - initial results
Bloody Hell, I knew that canon is happy to change lens protocols between bodies to lock out 3rd party users but doing it on the same model by tweaking firmware is just scary!
p.8 #5 · Contax N to Canon conversion - initial results
My Zeiss VS 24-85 with the Conurus conversion seems to be working fine on my 5D. I had one glitch over 2 days of shooting this weekend, where the camera just stopped working. I got error code 99 (I think). Turning the camera one and off erased thecode but the camera would then not operate or fire. (Caused me to miss a shot I was waiting for). Later, removing the battery and re-installing it seemed to get it working again. However, I don't know if this ia the fault of the camera or the lens The firmware in the camera is 1.1.0. Here is a sample with the VS 24-85: http://gallery.photo.net/photo/4958384-lg.jpg
p.8 #6 · Contax N to Canon conversion - initial results
Although it is nice to know that the first unit I shipped is still functioning (mostly), an occasional glitch like this still have to be fixed. I think when I am done with the reliability improvements, I will have to update all units out there regardless of whether they report problems or not, as a preventive measure.
The new version has these features:
1, a modified circuitry,
2, a stiffened construction which eliminates circuit board flex,
3, firmware improvements (still in progress),
4, "aging" test (don't ship working units right away but retest them after a few days), and
5, testing on more camera bodies and lenses and complete record of what is being tested (for the first time, a serial number will be written on each mount for tracking purposes).
p.8 #7 · Contax N to Canon conversion - initial results
Thanks, Bo-Ming, great that you are working so diligently on improving the kit, and keeping us informed. I'm really looking forward to trying it out. Keep going!
p.8 #8 · Contax N to Canon conversion - initial results
If it works on one camera and not the next, I suspect it's the start-up state of the 5D. For example, if the previous lens was a Canon 35mm F1.4, then the camera was turned off, a new lens mounted (say a F4 lens), then the operating system has to detect the change, query the aperture, etc. Just a guess, but it's probably a simple problem like that.
I look forward to the development of the 17-35 and 85.
Sep 18, 2006 at 06:24 PM
brainiac Offline [X]
p.8 #9 · Contax N to Canon conversion - initial results
I got error code 99 (I think). Turning the camera one and off erased thecode but the camera would then not operate or fire. (Caused me to miss a shot I was waiting for). Later, removing the battery and re-installing it seemed to get it working again.
Precisely that has happened to me a few times when not using a Conurus-N lens. Last time was at a wedding last weekend.
Lotus - the colours in the trees in that shot look rich and accurate. I can't wait to get this lens working.
I am beginning to worry that there is something wrong with my lens rather than bo-ming's circuitry even though I bought the lens new two weeks ago. I am going to put back the N mount and test it on an N body. However, according to the troubleshooting guide on conurus.com, an aperture of '00' means the camera isn't speaking properly to the conurus circuit. That should have nothing to do with the lens.
p.8 #10 · Contax N to Canon conversion - initial results
I found that the display varies from camera to camera. Perhaps 5D displays 00 even if the camera communicates with the circuit but the circuit does not communicate with the lens. Certain pins of the lens would display 00 on some camera models.
One of the upcoming firmware improvements is to make sure we display something wierd on more camera models, so long as we have a connection between the camera and the circuit. There will also be a reason code to indicate the fault (e.g. flashing the numbers 20, 00 and 5.6 on the display means a certain kind of communication failure.)
p.8 #12 · Contax N to Canon conversion - initial results
Bo-Ming,
I am glad to see you are working hard on the project and are making progress. Like herion, I too wait for the conversion kit ready to send my way. Despite Lotusm50 and brainiac's lens have glitches but I wish I have one with glitches right now.
p.8 #14 · Contax N to Canon conversion - initial results
In the image of the runner in the lake that I posted earlier today, there is some purple fringing around the runner's white t-shirt. (See crop below). Is this purple fringing due to the sensor or to some CA from the lens. Interestingly, in the post preceding this one with the sweaty runner, there is no purple fringing around the white t-shirt in that shot which was shot at virtually the same aperture (f8.0 vs. f9.0). http://gallery.photo.net/photo/4960775-lg.jpg
p.8 #15 · Contax N to Canon conversion - initial results
I thought purple fringing was also known as "sensor blooming", and so, due to the sensor, whereas CA usually shows as a red/cyan or blue/yellow shift. I'm hoping that's sensor blooming due to strong white/dark contrast, as it's pretty visible and could make this conversion less attractive for me if it's the lens. Then again, I haven't yet seen sensor blooming like that from a 5D, and hope it's not a 5D problem either.
Sep 18, 2006 at 09:48 PM
brainiac Offline [X]
p.8 #16 · Contax N to Canon conversion - initial results
Fantastic. That's why I use Zeiss. Look at the sweaty runner's hand. 3D. Like looking through a window. There is a sculptural sense of realism which the Zeiss lenses seem to capture more often than others. I hate to say it, but for me it is more exciting than the painting-like rendition of Leicas. I think its because of the contrast and very linear rendition of light and shade with high dynamic range. I think it's more important than colour, bokeh, and sharpness, because photography is about taking you there, and realistic light and shade is the one essential ingredient of the illusion. The enduring popularity of black and white argues this. Remind yourself by watching The Third Man again.
As regards the fringing, I have seen this fairly often when using my Zeiss glass on the 5D. I think it happens because the contrast is just so huge and the sensor can't handle the extreme highlight. You don't see it except around very strong highlights. Perhaps one day we will have sensors that can capture the full contrast range and confirm that it's not the lens' fault. Maybe when Canon catches up with Leica and goes 16bit. It doesn't worry me too much even if it is the lens' fault because everything else looks pretty good.
Here's an unaltered JPEG taken at night under street lighting on a 5D at iso 3200 with a Zeiss 35 f1.4. Despite the noise and circumstances, the eyes still have that glassy contrasty 3D look. http://braine.com/d/donal/20060908_002120.jpg
The reason I sold my APO-Summicron 180 f2 is because it made eyes look flatter, less glassy, with the glints kind of under control. I was looking at a picture, not through a window.
p.8 #18 · Contax N to Canon conversion - initial results
Looks very promising.
Sep 20, 2006 at 02:24 AM
brainiac Offline [X]
p.8 #19 · Contax N to Canon conversion - initial results
Update: I converted my 24-85 back to N-mount and tested on an N body. All worked fine, which suggests that there is nothing wrong with my lens. This all suggests that Bo-ming needs to improve the firmware on the adapter as it seems to work for some but not others. I have tried two. Both exhibited the same problem and could not take pictures.
p.8 #20 · Contax N to Canon conversion - initial results
I have to solve this mystery somehow, because I promised that I would solve all open customer cases before I would resume sales. Here are the data points:
1, you have two 5Ds, with firmware v1.0.3 and v1.0.5, and neither communicated with your N24-85 for more than 1 second.
2, you had your N24-85 tested on a Contax N body.
3, I've sent you two mounts, and neither worked for more than a second, and the second one I sent to you had a modified circuitry which improved reliability.
4, Lotusm50 had a 5D with firmware 1.1.0 and has been using his N24-85 for a few weeks. He had one intermittent glitch, subsequently fixed by removing the battery and reinserting it, probably because of the old circuitry, but nothing as serious as your problem.
5, at least one other customer has reported success with the 5D/N24-85 combination.
I would be shocked if the firmware version of 5D really mattered - but given the data so far is contradictory, I would have to stick with the original plan communicated to you earlier by email:I would make the latest version of the mount (not only with the modified circuit, but with the latest firmware and stiffened construction), and preinstall it on my N24-85. Then, I would beg/borrow/steal a 5D locally and make sure that works before I would ship the whole modified lens to you. It either works with your two 5D's or doesn't work.
1, if it does, your two 5D's are off the hook. You would then transplant the mount to your own N24-85 to see if that still works.
1a, If it still does, we are done!
1b, If it doesn't, you could send your lens to me so that I could see if it is an installation issue.
2, if it doesn't, maybe the firmware version of 5D's really matter, and maybe different production runs of 5D's behave differently.
