Topic: How to take great pictures of our glass?

[colink]

Has anyone got any tips/advice as to how to achieve great pictures of glass, my collection consists of 100% carnival glass and is quite difficult to achieve great pictures of it. I currently have a Sony A230 camera but im willing to purchase items where neccessary. Any help gratefully received ta  :thup: :thup:

[TxSilver]

Lighting is most important for good pictures. You have to decide whether your glass looks better in fluorescent, incandescent, or natural light. I haven't photographed but one carnival piece, so I don't know which is best. Your marigold or blue carnivals do best on a white background. White glass photographs better on dark blue or black. Unless it is a huge piece, use the macro setting (flower image). I simply point, focus, & shoot. I take several digital photos of each position, then save only the best one on the computer.

The next step is to take your picture to a good OCR program. I use PhotoDraw -- an old Microsoft product. Others use PhotoShop. There are many others that are pretty good. Some packages are free downloads. I only do four things with the OCR program if needed -- crop, adjust white balance, sharpen, and add a touch of contrast. The last thing (contrast) I do very conservatively, because it can change the appearance of the glass completely. (I usually end up doing one more thing -- cloning to get the stray rabbit hair off the background. I never see them until I'm processing pictures.)

Taking good glass pictures is not difficult if you can get the lighting right and know how to do the final processing steps.

[ahremck]

Just to add a titch to Anita's response.  You need to learn how to resize your wonderful piccies so you can post them for all of us to see.  It is best to get the sizing as close as you can to the 125Kb limit - 120+ is a good result.  I find the best way is to make multiple copies then experiment.  You will find that after a whilke you will guess very close first time.  The actual pixel count can vary enormously so dont get worried by that - it seems that glass with large areas of plain colour can be recorded with bigger pixel counts, but speckly, bust colours will tend to be smaller.

Ross

By pixel count I refer to that 534 x 631 for the Caithness Orchid vase below, and the 545 x 962 for the Whitefriars Kingfisher Blue vase - yet both are approx. 123 Kb

[KevinH]

Pixel count is really the number of pixels in height multiplied by number of pixels in width. The kilobte (Kb) value is derived from the format of the photo file (type = .jpg for most purposes within the GMB) and the amount of detail in the image.

Large areas of the same colour are recorded in a jpg file using an algorithm that minimizes the kilobytes used to store the information. What happens is that the algiorithm says "use the same colour value for all pixles in the stated area". But areas of high detail require much more kilobyte storage space as every part with differening colour has to be recorded separately in the file.

So, an image of say 600 x 400 pixels with a lot of detail will have a higher Kb size than a 600 x 400 image with much less detail. The Kb size of a file for given height / width is normally reduced through a process called "optimizing". This uses a very clever algorithm which effectively "throws away" unrequired pixel data whist retaining a seemingly ok picture (and it works because of the way the human eye and brain interpret colour and detail).

Although resizing the height and width of an image will reduce the Kb size it can take many attempts (as Ross has said) to get a Kb size below the 125 limit for the Board. It is the optimizing process that really does the best job. But many photo editing applications do not call it "optimizing" - preferring instead to talk about "image quality". Check your photo editor Help details for information on picture "qualtiy" ... it should be covered.

[TxSilver]

Have to put in a correction to what I wrote earlier. The photo editors are not OCR software packages. 

[Wuff]

Sorry, if this sounds a bit technical .

What jpg does to compress images (i.e. reduce file size) is to convert the image into frequencies, and then cut off the high frequencies (depending on the amount of compression). This is a bit similar to listening to music: if you damp the high frequencies, you will still recognise the piece, but it doesn't sound like a good recording. High frequencies in the "image" are needed to describe sharp contours - consequently a high compression ratio ("low quality") will affect regions with contours, but have little effect on large areas in the same colour. For the same reason, an image of identical pixel size and compression ratio will result in a smaller file if large areas without contour (like a uniform background) are in the image.

When working with jpg, keep in mind that even with the highest quality there is some compression (though usually unnoticeable in the image), i.e. loss of information. Each step of storing the image again, will result in additional loss of information (or image quality). If you are planning on doing other manipulations in the future, it is therefore recommended to store the "original" in a loss-free format (like .tif) - which will require more space on your disk, of course. If your camera only gives you jpg in the first place - make sure you keep this "original" for the future - never overwrite it with manipulated versions.

Below is an example of a very simple image - left stored as "quality 10", and right stored as "quality 0".

[colink]

Many thanks for all the tips/advice