Jump to content

Stargazers Lounge Uses Cookies

Like most websites, SGL uses cookies in order to deliver a secure, personalised service, to provide social media functions and to analyse our traffic. Continued use of SGL indicates your acceptance of our cookie policy.



  • Content Count

  • Joined

  • Last visited

Community Reputation

23 Excellent

About AndyEllis

  • Rank

Profile Information

  • Location
    High Spen
  1. I think that put simply, my point is that looking at any single factor is fraught with pitfalls - this applies equally to noise. A switched off sensor generates no noise at all but is clearly no good for imaging. A better, but still by no means perfect measure is signal to noise ratio (SNR). Perhaps it's best to say that a lot of features that Canon has built in to it's sensors in recent years have improved the signal and reduced the noise though not at every iteration - some provide a big increase in signal but have a marked increase in noise but still represent an increase in SNR, so the 450D is less noisy than the 500D, but the 500D is more sensitive and will produce better images. The step to the 550D increased signal and reduced noise so it is iteratively better than the 500D. The developments in the 600D focus on the flip screen and the sensor is very similar - the 650D uses the Digic 5 processor and is capable of a higher maximum ISO setting (25,600 vs 12,800) and has on sensor focusing, but otherwise again has no marked improvement in the sensor. In general terms, if you want one factor to look at, SNR is a good rule of thumb but there are still other factors to consider. In terms of the 1000D to 1100D step, the technology in the 1000D sensor relates most closely to the 450D and that in the 1100D most closely to the 550D so the step is larger. As I said before, the 1000D is a great camera, but the 1100D is markedly better in almost any metric.
  2. I meant 'cite' not 'site' - by that point in trying to explain it, my brain was going in to 'forced override standby mode'.....
  3. I was thinking it wasn't so I decided to add a bit more science... Well Capacity is related to dynamic range in that a greater well capacity leads to a greater dynamic range which is a measure of how well a sensor responds to both high and low levels of light, i.e., a greater dynamic range gives a more detailed signal across all values of incident photon density. On the surface, this may seem to be a good thing, but there are other considerations to take in to account here. Another factor that should be considered is Conversion Gain which is a measure of voltage change per single charge. Conversion Gain is also related to Well Capacity in that with a lower well depth, the change in the output voltage (which is related to charge divided by capacitance) is greater per single charge. It must also be remembered that as this is a digital system, read-out bit depth effects how the charge is converted to a signal at the ADC (Analogue to Digital Converter) stage. To put this in terms that can easily be understood, if you take a well of water that, when full can be emptied by using 8 buckets, if you then make 16 smaller buckets, the difference one cup of water makes will be more easily recorded as a change, i.e., if each of the smaller buckets can be filled by 5 cups of water and the larger by 10 cups, you only need to take away 5 cups of water before a change is recorded with 16 buckets, but you'd have to remove 10 before a change is recorded with 8 buckets. So, if there is a greater conversion gain at lower well capacity, a sensor with a lower well capacity given the same number of readout bits will better show low levels of incident photons but will 'fill' more quickly with high levels of incident photons and have a lower dynamic range. In terms of astrophotography, this means a lower well depth will more easily show changes in very low levels of light but a sensor with a greater well depth has the potential to give a smoother image. Above all of this and a more important factor is image bit depth, i.e., if you have a low bit depth, i.e., 8 bit, then a lower well capacity gives greater low light detail, so a good sensor for astrophotography should balance bit depth and well capacity or put in the most simplistic terms, a high well capacity is only a good thing with a high bit depth and generally speaking a balance of the two is what is achieved by manufacturers, i.e., a medium well capacity and medium bit depth. As I said before though, there are a lot of manufacturing factors to take in to account as well - pixel fill - there are electronics other than the sensitive material in each pixel - ideally, the sensitive material is presented at the face of the sensor and the electronics behind it, but this is difficult to do so only a proportion of each pixel's surface is sensitive to light. Over the years this has improved with manufacturing techniques, and with the introduction of micro-lensing, is beginning to approach a point where a far greater percentage of incident photons effect a charge change in the substrate with 100% as the ideal. Still all of this is theoretical and each different metric can not be taken alone outside the over all system and to latch on to one particular value such as well capacity and say that a greater well capacity gives a better picture is as erroneous as saying that the greater the number of pixels, the better. I hope this pleases the more scientific amongst you - if you like I can site references to some very technical papers on the subject. Andy
  4. The 1100D has a better sensor than the 1000D - it has micro-lensing which was introduced with the 1100D and 550D. The well depth, like noise, is a bit mis-leading when considering what the camera is capable of - it is a measure of the full capacity of each individual pixel well and though it is proportional to dynamic range (a better dynamic range gives better performance across different lighting levels) more important factors are Quantum Efficiency (which relates to how many photons are converted to electron hole pairs) and physical factors such as pixel size, pixel material, micro-lensing etc., are more important considerations, especially when considering low light performance. In the end analysis, the best way to make sense of it is to compare outputs, i.e., pictures, and the 1100D produces better pictures more easily than the 1000D though the 1000D is a good performer. There are albums from a 1000D user and an 1100D user here - http://www.facebook.com/astronomiser/photos_albums - these two images show the same target http://sphotos-b.ak.fbcdn.net/hphotos-ak-prn1/66454_304172279686469_1171849306_n.jpg (1100D) http://sphotos-f.ak.fbcdn.net/hphotos-ak-ash4/484816_304590756311288_545700511_n.jpg (1000D) By itself, this is purely anecdotal as the number of images stacked, the kit used, processing techniques are all different, but having seen hundreds of pictures from both cameras, it demonstrates adequately the differences I've personally noted and I have owned and used both cameras myself, though in my case, the weakest link was the guy behind the camera. Andy
  5. I imagine it'll work well with a firefly on the moon. I used one with an 098BL chipped webcam (Vesta Pro) on Mars and got some reasonable results with my 7.1" Mak Newt.
  6. Just about any dSLR where you can set the ISO to 400 or 800 and take timed exposures over 10 secs will do it - you might be out of luck with video, though you could set the camera to take a series of 10 sec exposures at ISO 800 (as an example) then make an avi out of them.
  7. Both types of sensor in the 4k are CCD, just that the sony one is an order of magnitude more sensitive
  8. Gary's article is good, but you must bear in mind that it is a study of noise and not of sensitivity - to put that in context, let's say a 450D has a noise value of 1 and sensitivity of 10 - yes, a 500D may then have a noise value of 2, but it is also more sensitive say 13. On the same scale a 550D would be 3 and 15. If you consider the gap, a 450D is '9' a 500D '11' and a 550D '12'. These are just arbitrary values, but the point stands. The 550D has improved pixel fill and micro-lensing (a tiny lens over each pixel to focus incident light on the sensitive material). In essence it is true to say that as the models have been developed, they have been better than previous models. Sometimes the step is small, but sometimes it is more noticeable such as the step between the 500D and 550D or the 400D to the 450D. Another question is ISO. ISO is handled by adjusting the substrate voltage and gain values in the sensor - steps between ISOs may involve both a voltage change and a gain change or either. This shows when you look at the noise generated at particular ISOs, i.e., ISO 400 may be very low noise ISO 800 high noise and ISO 1600 less noisy than 800... As always, every model is different and tests of different cameras of the same model also give varying results.
  9. There were two different sensors fitted to QC4Ks - this page shows my QC3K/ QC4K mod and at the bottom of the page, there is a pic showing the two different sensors - it must have the Sony ICX098BQ fitted for the mod to work. The CCTV camera may be a 1004x type, in which case John Groves' mod applies - http://jaggedplanet.com/ExViewMod.html
  10. Yes I saw that - the 350D's a great camera - I'm sure you'll get some great results with it
  11. Hap - that's not true - I've done over 200 1000D filter removals and have never had any cameras having a problem reaching infinity focus - including the one which I owned myself for two years and tried with dozens of lenses.
  12. I can see that this discussion is pretty much done with, but just out of interest, the 1100D vastly out-performs the 1000D and 350D. In the last few iterations of Canon's EOS cameras, they have introduced several improvements in the sensor including increased purity of pixel sensitive material, increased pixel fill and micro-lensing. All of this adds up to the fact that although the 1100D has smaller pixels than say the 350D and is noisier than the 450D, it is far more sensitive than either and capable of producing superior results. This image shows web-compression artefacts, but you can see how sensitive it is. This was taken by a guy that has being dSLR photography for 4 months. http://sphotos-f.ak.fbcdn.net/hphotos-ak-ash3/558950_287411964695834_841509111_n.jpg
  13. Beautiful image - is the 39th of November a way of denying the existence of December? clever... I'll have to say to the kids, 'No I haven't got you anything, there's nothing special about the 55th of November...?'
  14. my obs pc runs xp - cracking for anything astro and very stable - way better than vista - I got an xp install disk for £17 btw....
  15. You can flash the cam to look like an SPC900NC using an SPC binary - the softs to do it are here - http://www.burri-web...98/soft/wcrmac/ This is the binary for an SPC900NC - change the file extension from .txt to .bin - it won't allow uploads of binary files. spc900nc.txt
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.